Back
[01:36:57] <Loetmichel> mornin'
[01:44:47] <Loetmichel> soooo... just talked to the boss: granted me the mainboard P6tWS pro and the I7 - 2,66ghz that htt been spare ... i will pay with overtime hrs... now i have just to sort out the 12GB ram issue... and then i have a fat new PC for @ home
[01:44:57] <mazafaka> Loetmichel:
http://www.youtube.com/watch?feature=player_embedded&v=je4qV7fsVvU&gl=RU
[01:46:20] <mazafaka> http://cs307304.userapi.com/v307304406/2797/sc3dEvddt68.jpg
[01:48:19] <mazafaka> Loetmichel: for what will you use such a powerful PC? I need one for FEM calculations and 3D drawing...
[01:48:32] <Loetmichel> gaming, wat else
[01:49:09] <Loetmichel> have even ordered a GT650TI graphics board and a 2tb HDD to go with it ;-)
[01:52:06] <mazafaka> you're already adult, how can you play on PC?
[01:53:49] <mazafaka> Loetmichel: I myself plan to buy and RC trophy car for about 750 USD, but I know it is waste of time, because I will not be capable to walk the dog playing this thing, my dog hardly becomes tired from my cycling...
[01:55:24] <Jymmm> mazafaka: Try riding a NON stationary bike, then the dog might =)
[01:56:18] <mazafaka> Jymmm: i ride like in a link above,
http://a.d-cd.net/681e6u-960.jpg
[01:56:33] <DJ9DJ> moin
[01:59:48] <Jymmm> mazafaka: It was a joke.
[02:02:26] <mazafaka> I know
[02:04:30] <DJ9DJ> good night Jymmm & mazafaka
[02:04:31] <mazafaka> I finally close to my idea of using bikes and such, but I do not want any action cams anymore. It was funny to use BMX with only front brake on ice, but usual MTB with low seat is even a better rig for having fun.
[02:06:56] <mazafaka> I am waiting to by such thingies to put it under the car and it will keep it from falling while I will work with the car. Then there will be some details for lifting of the suspension, and some reworking of the wheel arcs, then there will be a roll cage.
[02:07:15] <mazafaka> I am talking about a real car, not RC car :)/
[03:00:01] <Loetmichel> MarkusBec: because i have saved me my inner childhood ;-)
[03:01:43] <Jymmm> Loetmichel: too bad you lost your adulthood =)
[03:16:58] <Loetmichel> Jymmm: i never had one
[03:17:05] <Loetmichel> :-)
[03:17:23] <Loetmichel> ia am a 43 year old child ;-)
[03:18:08] <Loetmichel> .... and proud of it!
[04:18:58] <mrsun> using a countersink (90 degree) for milling 45 degree angles on the side of stuff ? :P
[06:00:33] <MattyMatt> mrsun, why not it it's sharp enough?
[07:14:03] <mrsun> http://www.ctctools.biz/servlet/the-630/45-DEGREE-INDEXABLE-CHAMFER/Detail or use one of those
[07:14:11] <mrsun> i dont get it why the mechmate people doesnt do that? :P
[07:14:15] <mrsun> insted of grinding it
[07:14:29] <mrsun> its not expensive either
[10:09:15] <JT-Shop> http://www.youtube.com/watch?v=5MtX5M0AQFs
[10:11:56] <tjb1> Watching Modern Marvels makes me realize how much money we waste...
[10:12:12] <tjb1> Pardoning turkeys, sending them to disney world and then to live out their lives in a special farm?
[10:20:10] <skunkworks> https://www.youtube.com/user/brusspup/videos?view=0
[10:22:47] <sliptonic> JT-Shop: Do you mind if I ask how you did the estop circuit on your bridgeport conversion?
[10:22:47] <skunkworks> https://www.youtube.com/watch?v=hP4ai7dZXIo
[10:23:52] <skunkworks> sliptonic: I used example 4 -
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Sample_HAL_And_ClassicLadder
[10:24:27] <skunkworks> You can do the same thing with the estop hal component..
[10:25:50] <sliptonic> Cool. That's the kind of page I've been hunting for. Thanks.
[10:32:32] <JT-Shop> sliptonic: that is on my todo list... I do have the estop connected on the CHNC
[10:33:19] <sliptonic> skunkworks: Is that the estop_latch component?
[10:36:55] <skunkworks> yes
[10:37:33] <skunkworks> I played with it when someone had a question - I seems to act the same as example 4 (so you would not have to use classic ladder)
[10:37:49] <skunkworks> I need classic ladder so it really didn't matter to me
[10:39:03] <sliptonic> I've waded into that before and probably will again but I'm trying to get as much working through pncconf as I can before I go there.
[10:40:54] <JT-Shop> sliptonic:
http://linuxcnc.org/index.php/english/forum/search?q=estop_latch&childforums=1
[10:41:16] <JT-Shop> I knew I remember doing that for a machine but it wasn't mine it is a customers machine
[10:42:07] <JT-Shop> tjb1: did you test out the all in one homing/limit thing?
[10:42:21] <tjb1> Not yet, im gonna go out and try that soon
[10:42:46] <skunkworks> ah - it was jt :)
[10:43:58] <tjb1> Im waiting for it to get a little warmer outside
[10:55:53] <tjb1> Well its 45 out now, I suppose I can go out :)
[11:08:04] <awallin> hmm this is aparrently quite new...
http://www.youtube.com/watch?feature=player_embedded&v=39jFRUYRYYw
[11:15:04] <tjb1> I wonder how many people died trying to find the rock in Devil's Pool that pretty much keeps you from getting sucked off the edge.
[11:17:07] <pcw_home> awallin: pretty sure thats LinuxCNC
[11:18:04] <awallin> the rapids aren't set to "wow" mode clearly... :)
[11:18:31] <cradek> well css works...
[11:19:31] <awallin> if they would sell their mill with iso30 taper I could consider it - I guess they offer linuxcnc on the mills too?
[11:20:50] <JT-Shop> sliptonic: I have the estop hooked up now :)
[11:21:33] <Aero-Tec> the lathe looks nice
[11:21:48] <Aero-Tec> but looks like it will be costly
[11:22:05] <Aero-Tec> any idea the target price for it?
[11:22:48] <cradek> seeing the tap cycle I'm 100% sure it's linuxcnc. kind of weird they don't say it.
[11:23:30] <awallin> this one has been advertized in digital machinist... would be nice also..
https://www.machinetoolswarehouse.com/xcart/catalog/MD001-CNC-p-16346.html
[11:23:34] <Aero-Tec> I know mach can not do rigid taping
[11:24:27] <archivist> mack fails at turned threads too :)
[11:25:27] <roh> heh.. toolchangers.. i was thinking about buying this one
[11:25:28] <cradek> outside turning here:
http://www.youtube.com/watch?v=A_PSg0p0oYA
[11:25:29] <roh> http://www.schultheis-shop.de/index.php?cat=c66_Werkzeugwechsler-Werkzeugwechsler.html
[11:25:40] <roh> any comments about something like that?
[11:25:59] <JT-Shop> they used a center drill for a spotting dril
[11:26:00] <roh> as in.. is it worth the money/any problems you could see with it?
[11:26:48] <awallin> roh: how does the gripper work on that one?
[11:27:21] <roh> awallin: you push the lever up to unlock, and use it downwards as usually on a drill press
[11:27:25] <cradek> http://www.tormach.com/blog/tormach-cnc-lathe-prototype-fall-2012-update/
[11:27:52] <roh> http://www.youtube.com/watch?v=q_0c2E3J4bg thats a video of it in use
[11:27:55] <sliptonic> JT-Shop: You're faster than I but I imagine you hand edited files. I'm trying to use pncconf.
[11:28:21] <JT-Shop> I just copied and pasted the code from the forum and changed the input to match :)
[11:28:38] <roh> awallin: its a exchange part for the quill
[11:28:44] <roh> if i translated that correctly..
[11:28:48] <awallin> roh: and how does the position of the tool stay at constant z when you push down the quill?
[11:29:29] <JT-Shop> sliptonic: what does pncconf do when you specificy an external estop input?
[11:29:29] <roh> awallin: for milling one fixes it in the 0 position
[11:30:00] <roh> but having a quill is really nice, even on a mill i must say...
[11:30:15] <skunkworks> http://www.cnczone.com/forums/showthread.php?p=1196407#post1196407
[11:30:55] <skunkworks> I am not saying it...
[11:31:32] <awallin> huh? they don't have rigid tapping on the mills??
[11:33:42] <skunkworks> mach through printer port...
[11:34:09] <skunkworks> they have setups for linuxcnc - but no spindle encoder yet.
[11:34:17] <skunkworks> iirc
[11:34:36] <sliptonic> JT-Shop:
http://pastebin.com/EtcUd9zG
[11:34:46] <skunkworks> rogge from tormach was on the dev channel discussing the lathe and stuff
[11:34:54] <skunkworks> few months ago
[11:36:52] <roh> wow. nice lathe... expecially the puller
[11:37:05] <roh> s/expecially/especially/g
[11:38:33] <skunkworks> http://psha.org.ru/irc/%23emc-devel/2012-08-15.html#16:07:40
[11:43:20] <cradek> that one guy in that conversation sounds just like me. I agree with everything he says.
[11:43:53] <tjb1_> hey JT-Shop, that isnt working. Telling me pin 'limit' does not exist
[11:45:45] <tjb1_> http://pastebin.com/qxKxTMqC
[11:46:27] <jthornton> typo
[11:47:34] <tjb1_> I found my fat fingers
[11:47:36] <tjb1_> :(
[11:49:37] <tjb1_> Its not picking up limits with that
[11:53:28] <pcw_home> Wheres the smart editor that says "no such symbol 'limit' in line 33
[11:54:42] <pcw_home> "
[11:57:44] <mrsun_> hmm i wonder how to make reliable corner joints for 40x80mm tubing ...
[11:57:52] <mrsun_> i want the table of my mill disassembleable =)
[11:58:03] <tjb1_> What command will tell me the address of a new parallel port? "laspci -v" is just telling me where the I/O ports are but I remember another command that told me the new par port was 778 I think
[12:01:53] <archivist> ew Lindsey books shutting down/retiring
http://blog.makezine.com/2012/11/12/lindsays-technical-books-ships-last-catalog/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+makezineonline+%28MAKE%29
[12:01:57] <pcw_home> Can't you just specify parallel port 0 and parallel port 1 and let Linux.LinuxCNC deal with the addresses?
[12:02:45] <tjb1> Wouldnt I need to add the second in the hal? with the 0x378 out
[12:03:51] <pcw_home> I have not done it but indexes (0,1,2,3) are supposed to work
[12:09:59] <IchGuckLive> hi all
[12:10:13] <IchGuckLive> Aero-Tec: i got the diles
[12:10:17] <IchGuckLive> files
[12:10:42] <Aero-Tec> cool
[12:10:48] <Aero-Tec> how do they look?
[12:10:56] <IchGuckLive> good
[12:11:01] <Aero-Tec> cool
[12:11:06] <Aero-Tec> thanks
[12:14:52] <IchGuckLive> what is on your pyvcp only the rpm
[12:24:30] <Aero-Tec> it is the standard screen
[12:24:48] <Aero-Tec> what ever the live cd does
[12:25:01] <Aero-Tec> I have not changed it any yet
[12:25:12] <sliptonic> pcw_home: can the estop_latch component be made to work through pncconf without editing the main .hal file?
[12:27:06] <andypugh> That would have to be a question for ChrisM I think.
[12:28:07] <sliptonic> Sorry. Not here? I'll check back.
[12:35:07] <pcw_home> Yeah thats a question for Chris M but I don't thinks Hes much of a IRCer
[12:35:09] <pcw_home> I would ask in the pncconf section of the forum
[12:58:24] <IchGuckLive> did you check Machine/Show HAL Configuration if all the parport pins are there
[13:09:16] <Dolence> hi!
[13:10:03] <Dolence> guys, anyone tried to control one of these cheap chinese VFDs via modbus?
[13:10:58] <IchGuckLive> Dolence: it works on the mesacard perfect
[13:11:22] <IchGuckLive> in standard VFD controll mode at the spindel interface from the 7i76
[13:12:48] <Dolence> whats a mesacard?
[13:14:27] <IchGuckLive> http://www.mesanet.com/
[13:14:55] <IchGuckLive> Dolence: woudt you like to do it from the parport
[13:16:48] <pcw_home> I know some drives can be driven by the modbus component (GS2?) but it might need some customization for your particular drive. Also I think there is a more generic modbus component that has just been added to Master
[13:16:51] <Dolence> i was thinking in using modbus (serial port)
[13:21:58] <pcw_home> I think the mb2hal driver should work with the right setup and assuming you have good enough specs. Heres an example mb2hal ini file:
[13:22:00] <pcw_home> http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=blob;f=src/hal/user_comps/mb2hal/mb2hal_HOWTO.ini;h=98103cd5b0f124eaa2ec085d2d72cee30aaa333d;hb=725822cac8f621c40c7c6105fb1a845eca335705
[13:24:55] <Dolence> what you mean by good enough specs? PC specs, like cpu clock and ram?
[13:28:38] <pcw_home> No good enough modbus register specs for the VFD to be able to control it
[13:29:25] <Dolence> I see, good documentation
[13:29:28] <pcw_home> (and proper modbus implementation, thats an issue sometime as well)
[13:30:30] <Dolence> so a dac to 0-10v interface could be a better solution?
[13:35:05] <tjb1> Oh goody, my torch fires when you remove power from the box!
[13:35:09] <tjb1> Just want I wanted!
[13:38:17] <pcw_home> Dolence I dont know, the modbus thing is worth a try if you dont mind using master
[13:40:49] <IchGuckLive> im off By
[13:46:18] <Dolence> pcw_home i think it's too much to deal for me, I could do it, but with my limited skills it would take a while
[13:46:55] <Dolence> maybe doing it via parallel port and a 0-10v interface could be a faster solution
[13:52:44] <pcw_home> You would be a trail blazer using modbus with a new drive, on the other hand it would probably help others with the same drive (and the guy that wrote the driver is often on the LinuxCNC forum so is likely available to help)
[13:57:38] <tjb1> Ive unhooked nearly everything, torch still starts when g540 loses power
[14:00:11] <kwallace> Dolence: Have you seen this yet? :
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?VFD_Modbus
[14:02:58] <pcw_home> You probably need to torch to be enabled by the chargepump
[14:10:36] <Dolence> kwallace yes, I was reading this doc right now... I will control vfd manually while I do some experiments with modbus
[14:10:40] <mevon> hi everyone, just wondering is it normal that when a limit switch is tripped, the whole machine goes offline?
[14:10:59] <mevon> have to hit f1+f2 again to get it back
[14:13:54] <pcw_home> I would guess that hitting the hard limits in normal operation is serious fault
[14:15:36] <kwallace> mevon: The hardware limits are considered an e-stop condition. Setting up the soft limits just inside of the hardware limits is the standard practice and keeps the machine in a normalish state, but you may need to click an override button to jog off the soft limit.
[14:19:06] <mevon> ok thanks ill consider that, but i cannot seem to find the over ride button
[14:20:25] <JT-Shop> sliptonic: I answered your question on the forum
[14:21:10] <kwallace> I recall it being close to the spindle control buttons in manual mode.
[14:21:23] <tjb1> JT-Shop: Gecko thinks my G540 is damaged, its activating outputs without charge pump, e-stop on, no db25 and it does this when power is removed from it
[14:22:02] <kwallace> The check box may not show up if your .ini file isn't set up for soft limits.
[14:24:13] <kwallace> tjb1: Just in case, there is a charge pump override jumper in the G540 which may keep it enabled whenever power is applied.
[14:28:00] <tjb1> kwallace: Its in fault mode, i unplug it and it activates the outputs
[14:29:13] <pcw_home> Sure sounds like the chargepump is busted (to use a technical term)
[14:30:49] <kwallace> tjb1: Okay, just checking. Also just in case:
http://wallacecompany.com/machine_shop/G540/
[14:31:18] <tjb1> kwallace: What am I looking for there?
[14:32:17] <tjb1> Ive looked all over it hoping I screwed up...
[14:32:31] <tjb1> One wire for arc start going to the output and the other going to the G540 power supply ground
[14:33:08] <tjb1> I can come into LinuxCNC and do M3 and turn it on and M5 to turn it off, I unplug power supply for G540 and the plasma cutter turns on
[14:33:18] <pcw_home> Are these OPTO coupler outputs?
[14:33:23] <kwallace> tjb1: These are just part of my notes on getting the charge pump working with a parallel port.
[14:33:54] <awallin> tjb1: that sounds like grounding-problems that optocoupling might solve..
[14:35:03] <tjb1> It worked before this exact same way
[14:35:18] <tjb1> Ill send it in next month once im done with my presentation
[14:36:11] <awallin> a design where the power-stage of something turns on when you cut the power to the control/logic part is not very sound...
[14:38:36] <tjb1> It has to be something I did…its turning on with another parallel port
[14:40:43] <pcw_home> I thought you said it still enabled the outputs at power off even with no DB25 from host
[14:43:26] <pcw_home> If thats the case (and the chargepump is enabled) Thats probably a bad G540
[14:43:55] <Aero-Tec> just to make sure I am getting things right
[14:44:21] <Aero-Tec> touch off from fixture keep the DRO 0ed
[14:44:35] <Aero-Tec> and is used for setting up tool table
[14:45:22] <Aero-Tec> and touch off from work is for setting the work offsets?
[14:45:29] <Aero-Tec> is that right?
[14:46:07] <Aero-Tec> still working with my lathe
[14:47:21] <tjb1> pcw_home, I have two wires for the plasma cutter…when touched the torch starts. I have one hooked to a ground that all power supples share and the other to the output. Does that sound correct?
[14:48:45] <Aero-Tec> would not a relay be better?
[14:49:08] <Aero-Tec> it would isolate the plasma from the CNC
[14:49:50] <pcw_home> Are either of the 2 wires from the plasma torch grounded?
[14:49:55] <Aero-Tec> plasma/ router will be the next thing made
[14:50:07] <abetusk_> um, open source eagle .brd to gerber file program?
[14:50:12] <tjb1> pcw_home: I ground one of them
[14:50:21] <abetusk_> or rather, any source on the eagle .brd or .sch file format?
[14:52:01] <Aero-Tec> tjb1: is the wire you grounded to the cnc a ground for the plasma?
[14:52:26] <pcw_home> But if you connect them to enable the plasma, one must have a power source (signal)
[14:52:27] <pcw_home> so is it possible you have signal and ground reversed?
[14:52:44] <tjb1> Ive reversed them…same thing
[14:52:58] <tjb1> Hypertherm doesnt say which wire is which…just 5 and 6 are arc start
[14:52:59] <Aero-Tec> you should disconnect the plasma wire grounded to the CNC and see if it is a plasma ground
[14:53:09] <pcw_home> are you using the G540 1Amp outputs?
[14:53:22] <tjb1> Im using the two outputs on the G540
[14:53:26] <tjb1> Ive tried both outputs
[14:53:39] <tjb1> Aero-Tec: The plasma ground isnt hooked to the table at the moment, it did it with it connected also
[14:54:10] <pcw_home> the G540 has high current outputs and TTL outputs, which are you using?
[14:54:47] <tjb1> I only see 2 outputs
[14:54:56] <tjb1> Pin 17 and 1
[14:55:28] <Aero-Tec> why not a relay?
[14:55:56] <tjb1> Why should it need a relay?
[14:56:14] <Aero-Tec> would that not be the best and safest?
[14:57:00] <tjb1> It worked before fine
[14:58:43] <Aero-Tec> when I said ground I meant DC power ground and not ac power in ground
[14:59:02] <pcw_home> does it just turn on briefly when you pull the G540 plug or is it stuck on
[14:59:11] <pcw_home> ?
[14:59:54] <Aero-Tec> I would measure the volts from the plasma wires you have to the DC ground of the plasma
[15:00:14] <pcw_home> Yep this can be sorted out with a voltmemter
[15:00:18] <Aero-Tec> see what one is 0 or if they both have voltage
[15:00:44] <pcw_home> voltmeter
[15:02:04] <Aero-Tec> calling the lathe gurus
[15:02:20] <tjb1> I hooked relay up
[15:02:38] <JT-Shop> from the hypertherm manual start plasma requires dry contact closure to activate...
[15:02:40] <tjb1> One coil wire to 12v, other coil wire to output of the G540. Ground for 12v shared to g540 ground
[15:03:11] <tjb1> NO contacts, one with 1 wire to plasma, the other with the other wire
[15:03:29] <tjb1> Plug 12v power in without G540 power, arc starts…turn g540 on and relay turns off
[15:03:48] <Aero-Tec> just wanting to check on setting things up tool table, when to use the fixture touch off and when to use the work touch off
[15:04:07] <pcw_home> sounds like the G540 is bad
[15:04:40] <JT-Shop> tjb1: I never turn on my plasma until Linuxcnc is in control
[15:05:01] <JT-Shop> or I get a pffffit from the electrode
[15:05:49] * JT-Shop wanders off
[15:06:05] <tjb1> Guess I will grab one of these…current way of setting wiring up sucks -
http://www.ebay.com/itm/OMRON-MK2P-I-DC-12V-General-Purpose-Relay-10A-With-PF083A-Socket-Base-/290739764281?pt=LH_DefaultDomain_0&hash=item43b170bc39
[15:13:20] <Aero-Tec> that should do the trick
[15:15:53] <tjb1> Im serverly confused...
[15:15:53] <tjb1> I am using this BOB also
[15:15:53] <tjb1> http://cnc4pc.com/Tech_Docs/C10R10_WG.pdf
[15:16:06] <tjb1> When one side of coil is hooked to pin 17 in the outputs and the other is hooked to the GND, it activates relay no matter what...
[15:16:57] <tjb1> When it activates relay….pin 17 to ground shows infinite resistance
[15:17:24] <tjb1> What magic is happening...
[15:17:41] <Dolence> guys, I recently switched my old rotary tool to a spindle+vfd, now, everytime the spindle starts to spin my axis moves a little bit... I tried to ground everything
[15:17:55] <Dolence> but still giving me some kind of noise problem
[15:19:16] <Dolence> everything is grounded to a common point
[15:22:36] <kwallace> Do you have a filter on your VFD power input? It's the shiny bit here:
http://www.wallacecompany.com/cnc_lathe/HNC/00024-1a.jpg . Every VFD I've installed (not that many) has needed one.
[15:24:14] <pcw_home> VFDs put ~360V squarewaves into the motor (which has maybe 500 pF capacitance to GND)
[15:24:16] <pcw_home> havoc ensues.....
[15:25:05] <tjb1> c10 outputs are different...
[15:25:06] <pcw_home> a common mode choke (giant ferrite bead) on the UVW wires helps a lot as well
[15:26:53] <mrsun> shielded cables etc?
[15:28:09] <pcw_home> shields may help but the main issue with VFDs is the unlimited current pumped into GND via the fast rise times into the motor/gnd capacitance
[15:30:06] <Dolence> cables between vfd and mtor are not shielded
[15:33:35] <pcw_home> Good frame GND on VFD is important too
[15:35:01] <pcw_home> so the high frequency return current loop between the VFD and motor is as short as possible
[15:35:11] <Dolence> solved!
[15:35:24] <Dolence> serial port ground wasnt grounded on driver
[15:35:34] <Dolence> it's an old driver, bad stuff
[16:02:30] <tjb1> Can you type multiple lines on one line in axis mdi and have them do it
[16:02:34] <tjb1> Like
[16:02:41] <tjb1> "M3 S1, p4, m5"
[16:07:00] <sliptonic> JT-Shop: Thanks. Estop working now.
[16:13:53] <andypugh> abetusk: You might want to hope that djdelorie sees your question
[16:14:10] <djdelorie> what?
[16:14:28] * djdelorie checks logs...
[16:15:49] <andypugh> Hmm, actually, I don't know why you sprang to mind. I got confused between Eagle and GEDA
[16:15:59] <djdelorie> I don't use eagle, so I don't know... there might have been some efforts in that direction, but certainly, the easiest way to get gerbers out of eagle is to run eagle :-)
[16:16:16] <djdelorie> we do have a desire to migrate eagle design files to geda though
[16:19:39] <andypugh> JT-Shop: You recall that dividing head I was looking at on eBay? I just got back from collecting it.
[16:20:01] <andypugh> It seems rather solid.
[16:20:24] <archivist> where what....
[16:21:20] <mrsun> will have to go and buy materials on monday =)
[16:21:22] <mrsun> yeey
[16:21:24] <mrsun> expensive :P
[16:21:25] <andypugh> One of these:
http://www.chronos.ltd.uk/acatalog/Engineering_Menu_Vertex_Dividing_Heads_90.html
[16:22:24] <pfred1> I have to give geda another try
[16:22:30] <archivist> I have a vertex, but not that model
[16:22:57] <mrsun> hmm how to measure long lengths to quite accurate distances .. :/
[16:23:11] <andypugh> Taper-roller bearing spindle sounds like a great advance on my nasty rotary table
[16:23:25] <archivist> but only 40-1
[16:23:40] <andypugh> I will put a servo on it anyway.
[16:23:56] <andypugh> mrsun: laser interferometry
[16:24:01] <mrsun> ough
[16:24:28] <andypugh> laser time-of-flight is good for distances over 100m or so.
[16:24:37] <archivist> andypugh, watch for backlash, I only rotate one direction normally
[16:24:44] <mrsun> andypugh, ok not that long
[16:24:48] <mrsun> was thinking more 1.5 meters :P
[16:25:11] <archivist> lots of J blocks
[16:25:44] <andypugh> mrsun:
http://www.ebay.co.uk/itm/190744612617
[16:26:30] <mrsun> heh =)
[16:26:40] <archivist> I have a vernier does few feet
[16:26:57] <andypugh> What's the budget?
[16:27:02] <mrsun> no budget :(
[16:27:09] <DJ9DJ> gn8
[16:27:10] <mrsun> just want my tubing cut to the right length :P
[16:27:28] <archivist> roller and encoder
[16:27:39] <andypugh> A good tape-measure will get you 0.05%
[16:28:40] <archivist> cnc band saws have encoders, grip metal and move it with a leadscrew
[16:29:10] <andypugh> Marking _on_ the lines is always more accurate. Bear in mind that a conventional vernier works on the principle that you can spot a misalignment of 0.05mm with the naked eye.
[16:29:35] <andypugh> There is no magic in a vernier, just a line everywhere you need one.
[16:30:30] <archivist> I must take pics of this vernier... it is different, has pegs in holes, uses the hole spacing difference in the slider v frame
[16:31:11] <archivist> so digital vernier to 4 ft at every 1/4 inch
[16:34:50] <mrsun> i guess that the tubing length is the least problem tho ... corner joints without welding will be a fun one also :P
[16:37:46] <andypugh> Aero-Tec: Typically I touch-off the "master" tool (my favourite turning and facing tool) into G54, and then touch-off the other tools into the tool-table. That means that the tool-table contains the offsets from Tool1.
[16:38:14] <andypugh> Then, if I touch-off any other tool to the work into G54, everything else moves (correctly).
[17:16:23] <tjb1> Bad news for the arduino sampling torch voltage…its picking up the noise also
[17:16:29] <tjb1> so I am going to try to average it
[17:20:46] <roh> tjb1: can#t you filter it out?
[17:21:07] <tjb1> Im going to try to average the last 5 or more inputs
[17:21:57] <roh> try putting some resistor in series with your adc line and put some small cap from the adc in to gnd
[17:22:17] <tjb1> adc?
[17:22:30] <roh> analog in of the arduino
[17:23:04] <roh> so.. signal -> resistor (10k) -> one side of the cap and arduino input
[17:23:39] <roh> and gnd -> other side of the cap. 10nF should be enough
[17:23:58] <roh> how fast/often do you sample?
[17:24:05] <tjb1> Doubt I have a 10k resistor or a 10nf cap
[17:24:07] <tjb1> as fast as arduino runs
[17:24:28] <roh> how fast do you need it to react? a low-pass obviously slows the slope
[17:25:27] <tjb1> decently fast
[17:28:20] <tjb1> lets see if I remember how to check caps with my multimeter...
[17:29:35] <tjb1> all I have is .1 and 1 uF
[17:30:59] <andypugh> ten 0.1 in series ?
[17:31:20] <PCW> 1K and 1uF is 1 ms time constant which should be plenty fast
[17:31:20] <tjb1> I got like 3 of them
[17:32:05] <andypugh> 1uF SMT 0804. What voltage are those likely to be? Presumably very low?
[17:32:34] <tjb1> My cap says 1uF 50v
[17:32:57] <tjb1> So I can do the above with 1uF and 1k?
[17:33:43] <roh> try it out
[17:35:36] <PCW> 1 mS time constant 159Hz bandwidth should be OK
[17:36:17] <tjb1> So voltage divider output -> resistor -> one leg of cap -> same leg of cap to analog input
[17:36:21] <tjb1> other leg of cap to ground
[17:40:03] <tjb1> I cant get it to print the real damn voltage…hold on
[17:40:10] <tjb1> Too much math for arduino >:(
[17:40:40] <tjb1> Too hard for little arduino - realVoltage = (((tip*(5/1023)*(7000))/5000)*50);
[17:41:06] <tjb1> It doesnt like the 5/1023 or the .004887585
[17:41:57] <andypugh> 5.0 / 1023.0
[17:42:58] <andypugh> What you have will be done as integer maths, and the first 5/1023 will be zero.
[17:52:42] <tjb1> This list looks pretty good
[17:52:43] <tjb1> http://pastebin.com/qmhMVTKK
[17:52:52] <tjb1> Yes the top voltage is really the low…the torch needs moved up quite a bit
[17:53:00] <tjb1> top = tip
[17:53:29] <tjb1> Thats also using a software average of 15 values
[17:55:08] <tjb1> Im guessing the ~40 values are values stuck in the list from the high pierce
[18:01:14] <L84Supper> I just got back from www.automationtechnologiesinc.com (aka Keling) the guy that runs it has Phd in Mech Eng for the UIC
[18:06:39] <roh> tjb1: arduinos do fixed point math. no fpu in there
[18:07:23] <s1dev> roh: I haven't seen any of the conversation at all but floating point can be implemented in software
[18:07:39] <roh> also.. what resistor values is your voltage divider? depending how that one works you will need no or a different resistor in front of the cap
[18:08:15] <roh> s1dev: sure. but on an avr nobody does that. you simply explain the little critter your problem in fixed point math
[18:08:16] <tjb1> 2k/5k
[18:08:28] <s1dev> anyone have an opinion of the small benchtop Grizzly lathes?
[18:08:31] <s1dev> roh true lol
[18:08:39] <roh> tjb1: so you measure something like 7V max?
[18:08:43] <tjb1> yes
[18:08:59] <tjb1> tip voltage was not 30…its 90…arduino isnt getting something right
[18:09:33] <andypugh> tjb1: You need to change the order then, do the multiplies first. That 5/1023 will make the result always zero
[18:09:41] <roh> well.. i dunno what internal resistance an avr has... but the 1k shouldn load the circuit that much
[18:09:57] <tjb1> andypugh: Its not even printing the first value correctly
[18:10:31] <jdh> do the a/d counts look reasonable?
[18:11:08] <tjb1> It should be 263 out of 1023 and its giving me like 50-100
[18:11:41] <tjb1> If my math is right…1.8 in before voltage divider, 1.286 coming out of divider into arduino, 1.286/(5/1023)
[18:11:52] <roh> do something like "voltage in mV = adc * 6.83593"
[18:12:18] <roh> thats (7k/5k)*5000mV/1024
[18:12:23] <jdh> hook a battery up to i.
[18:12:26] <jdh> it
[18:13:04] <roh> measure before and after the 1k to ground and see if there is a difference
[18:13:17] <roh> there shouldnt be (much).. few mV max.
[18:13:28] <tjb1> I took that filter off too
[18:13:33] <tjb1> and it still isnt giving me correct values
[18:14:06] <roh> can you pastebin your code?
[18:14:58] <tjb1> http://pastebin.com/q270XcQv
[18:16:41] <ReadError> L84Supper: oh word
[18:16:46] <ReadError> thats where i get all my stuff from
[18:16:47] <ReadError> mostly
[18:16:52] <ReadError> how is their setup?
[18:17:18] <L84Supper> ReadError: just a warehouse with an office in front
[18:17:34] <PCW> 2K and 5K resistors swapped? (that would make 2.5 X error)
[18:17:50] <ReadError> L84Supper: did you see their printers?
[18:17:51] <ReadError> the 3d printer
[18:17:55] <L84Supper> he selling an FDM printer now, open source friendly
[18:17:56] <roh> i dont understand these loops in there...
[18:17:56] <tjb1> im measuring it with my multimeter, im getting 1.270 out of the divider
[18:17:59] <ReadError> ya
[18:18:07] <ReadError> thats what i was curious about ;)
[18:18:09] <tjb1> roh: Its building a list 15 long and averaging it
[18:19:06] <L84Supper> ReadError: he sells them for the makerslide guy
[18:19:19] <roh> hm.. you dont need a list for that ;)
[18:19:27] <L84Supper> he was working on some software for it when I came in today
[18:19:28] <tjb1> I didnt write that code
[18:19:28] <ReadError> did you get to see one L84Supper?
[18:19:54] <L84Supper> ReadError: yeah, it been around for a bit, he wants to carry more pro printers
[18:20:47] <L84Supper> http://www.automationtechnologiesinc.com/products-page/3d-printer/ord-bot-hadron-3d-printer-kits
[18:23:08] <L84Supper> also stopped by
http://www.shars.com/
[18:23:50] <L84Supper> that place is busy
[18:29:42] <r00t4rd3d> http://www.youtube.com/watch?v=rAeQn5QnyXo
[18:29:46] <r00t4rd3d> that is awesome
[18:29:52] <r00t4rd3d> damn thing can run
[18:29:52] <andypugh> L84Supper: That Hardon is quite a neat design.
[18:29:52] <roh> tjb1: try
http://pastebin.com/w9QijJEq
[18:30:20] <tjb1> roh: Its not giving me the correct value to begin with
[18:30:33] <roh> that one 'averages' 5 samples
[18:31:55] <roh> ah.. and swap all the doubles for int.
[18:32:33] <roh> well.. atleast the tip and last
[18:32:43] <roh> maybe its broken conversion to double
[18:32:51] <tjb1> im stupid
[18:33:04] <tjb1> the voltage divider was backwards…im sorry to whoever already mentioned that
[18:33:18] <tjb1> It was giving 1.4 v to arduino which is ~289
[18:33:49] <tjb1> ok im going to eat
[18:38:18] <andypugh> is this a forum glitch, or do we need to pay more attention to this?
http://imagebin.org/236212
[18:38:56] <abetusk> djdelorie, any effort on migrating eagle files to geda?
[18:39:17] <djdelorie> you'd have to ask on the geda-user mailing list, or #geda on OFTC
[18:39:57] <andypugh> <ponder> is abetusk in orbit round Mars?
[18:40:11] <djdelorie> we'll find out in a few hours :-)
[18:40:21] <abetusk> ?
[18:41:32] <L84Supper> http://www.raspberrypi.org/phpBB3/viewtopic.php?f=37&t=14122 LinuxCNC (Emc2) How to compile and get it working.........
[18:42:23] <abetusk> #geda redirects you to #eda? or do you need to be invited to #geda?
[18:42:32] <djdelorie> on OFTC, not freenode
[18:42:39] <abetusk> ah
[18:43:00] <L84Supper> andypugh: IMHO the cubieboard is a better fit for LinuxCNC
[18:43:17] <L84Supper> http://cubieboard.org/
[18:43:58] <L84Supper> nobody has tested the GPIO's yet for IRQ timing or bit-bang speeds
[18:44:02] <andypugh> Yes, but Pi has a user base and momentum. Beta was better, VHS won
[18:44:44] <L84Supper> i think they will live side by side
[18:45:38] <L84Supper> the allwinner A10 is in tons of products
[18:45:51] <L84Supper> even $30 dongles
[18:47:00] <roh> pi hasnt any serious amount of gpio
[18:48:27] <L84Supper> you can also buy a10 chips now for ~$10 ea
[18:49:34] <L84Supper> anyone ever buy from
http://www.metalsupermarkets.com/MSC-home.aspx ?
[18:50:17] <L84Supper> skip the web orders if you do and just call the local store for better deals
[18:50:31] <andypugh> I have had metal from them, but collected by the guy who wanted the thing made.
[18:50:47] <andypugh> I buy a lot of metal from eBay
[18:51:14] <andypugh> Or Metals4U
[18:51:44] <andypugh> Or if I want a 3m length, and can borrow a car. from the nornal stockholders
[18:52:47] <L84Supper> 21 ft here, I forgot to bring a saw once and drove 10 miles with a full section o the top of a Jeep
[18:53:30] <andypugh> 7m is the other length, but that is "no way"
[18:53:45] <andypugh> L84Supper: Did you see my PM?
[18:57:14] <jdh> the "high speed wifi' in this motel is 52Kb/s
[18:59:22] <L84Supper> broadband prices at dialup speeds
[19:00:17] <andypugh> I rescued an acoustic-coupler terminal from the skip yesterday.
[19:00:33] <andypugh> Finding a phone to fit it might be hard
[19:00:35] <jdh> did you ship it to north florida?
[19:00:54] <andypugh> And it turns out to not be the relic I thought.
[19:01:42] <andypugh> In fact, if you want one, they are still on sale:
http://www.uic.edu/depts/accc/telecom/tdd.html
[19:02:32] <jdh> pricey
[19:06:27] <roh> L84Supper: nice board (cubie) .. lets see when they are available without waitinglists
[19:07:41] <andypugh> My jongoism makes me prefer Pi
[19:07:48] <andypugh> (jingoism)
[19:12:55] <L84Supper> roh; there's also the
https://www.miniand.com/products/Hackberry%20A10%20Developer%20Board
[19:13:02] <L84Supper> Hackberry
[19:13:28] <jdh> my replacement Pi ADC seems to have finally been delivered.
[19:13:41] <tjb1> Look at this output :D
http://pastebin.com/RQpwd8cS
[19:14:56] <tjb1> starting at line 70 its holding withing 2 volts
[19:15:26] <tjb1> That may be the plate isnt flat or the water table isnt completely level
[19:37:42] <tjb1> https://dl.dropbox.com/u/3090951/filtervsnot.jpg
[19:37:49] <tjb1> Red is unfiltered, green is with the filter
[19:51:05] <tom3p> tjb1 is that graph the gap voltage for a torch or similar?
[19:51:35] <tjb1> yes
[19:51:39] <tom3p> its very much the ionization/breakdown/discharge features of an EDM spark ( my work is all EDM) but is it voltage? & what are the vert & horz units?
[19:51:51] <tjb1> vertical is arc voltage on plasma
[19:52:05] <tjb1> horizontal is really nothing…i havent timed it yet so I just threw that in there
[19:53:04] <tjb1> Im plotting different filter settings
[19:53:42] <tom3p> teh 120 vertical is volts? i guessed its voltage, but 120V sensing is nuts, I use 21K to 1K res divider for 115V dc and get near 5v out
[19:54:11] <tom3p> (i dont mean what you did is nuts, i means its a bit dangerous )
[19:54:42] <tjb1> its the arc voltage
[19:54:52] <tjb1> Im really only measuring 0-5
[19:54:56] <tom3p> ok
[19:55:10] <tjb1> Plasma cutter steps 350 down to 7 and i step it down to 5
[19:55:47] <tjb1> just trying to get the best line :)
[19:56:07] <tom3p> gtocha, and the horz scale is 1 tic per sample? (whatever poll speed that is on an arduino?
[19:56:22] <tjb1> Yeah im not sure what it is, im trying to figure out how to add a timer here
[19:56:47] <tom3p> in EDM we filter out anything over 500hz ( besides the machines cannot respond to even the 500hz changes )
[19:57:31] <tom3p> got a real scope? you can calibrate and forgo the constant timing
[19:58:08] <tjb1> I do have a scope but im not too hot with it
[19:58:20] <jdh> add millis() to the write line
[19:59:08] <tom3p> it will usually have its own calibration knob/btn, and thats usually a 1khz 5V sqr wave, that oughtta tell you what the arc is doing
[19:59:08] <tjb1> Thats it?
[19:59:35] <tjb1> This is a plasma cutter, im not looking at the arc just the voltage
[19:59:36] <jdh> that will return the # of mS since boot
[19:59:37] <tjb1> :)
[19:59:57] <jdh> or, save the last value, subtract from the current value.
[19:59:59] <tjb1> jdh I want it to tell me how much time has passed from voltage > 150 to voltage < 150
http://pastebin.com/b8HqnVDp
[20:00:23] <jdh> you just sucking it in to excel?
[20:01:02] <tjb1> yes
[20:01:46] <jdh> Serial.print(millis());
[20:01:46] <jdh> Serial.print(",");
[20:01:59] <jdh> put that before your voltage print
[20:04:10] <jdh> or, save the previous millis() and subtract to just show elapsed time between samples.
[20:04:28] <tom3p> btw if you use a scope near any arc discharge, use a cheater on the ground lug of the cord, DONT have a grounded scope ( or invest in scope lead stock )
[20:05:13] <tom3p> and doesnt arduino have micros() as well?
[20:06:27] <jdh> yeah, but kind of overkill for this.
[20:06:34] <tom3p> the hi to low transistion ( when it overcomes the dielectric breakdown strength of air ) aint milliseconds wehn V drop current rises
[20:07:16] <tom3p> "tell me how much time has passed from voltage > 150 to voltage < 150" maybe i misunderstood
[20:07:48] <tjb1> voltage value 150 out of 0-1023
[20:07:50] <jdh> that's not particularly relevant information
[20:08:21] <jdh> (microseconds between )
[20:08:52] <tjb1> commas != new colum in excel :(
[20:09:13] <jdh> data, text-to-table
[20:09:38] <jdh> select the column, data menu
[20:10:10] <tom3p> it tells you dielectric breakdown time, and the current rise time, i dunno what he wants the info for
[20:10:19] <jdh> text-to-column maybe
[20:10:37] <tjb1> Got it
[20:10:44] <tjb1> tom3p this isnt an EDM
[20:10:49] <jdh> if the value is in microseconds, it isn't useful
[20:10:59] <tjb1> Im trying to get a smooth voltage input for a torch height controller
[20:11:11] <tjb1> Im testing which filter is going to give me that
[20:11:35] <jdh> you can change the code to print out every 100mS or wtf
[20:11:40] <tom3p> tjb1 it doesnt matter when you look at high voltage suddenly dropping to low voltage across a gap, its a discharge, good luck with your project
[20:12:06] <tjb1> That variation is the torch moving closer to the work which is resulting in the lower voltage :)
[20:12:19] <tjb1> The arduino isnt printing the value until it reaches a certain value so im not getting garbage
[20:13:01] <tjb1> The start is when the arc is formed, it peaks and then the torch moves from pierce height to cutting height and that is where the voltage stabilizes. It is taking longer than I would like because I am averaging 15 samples but oh well
[20:14:18] <tjb1> jdh, whatever I did in text to columns messed it up
[20:14:22] <tjb1> It broke it like half up
[20:19:09] <tjb1> jdh: The green here is 7 averaged samples with a 1660 ohm - 1uF filter and the red is unfiltered with 15 averages
[20:19:18] <tjb1> https://dl.dropbox.com/u/3090951/7samples.png
[20:19:50] <tjb1> Green is same in this one, purple is 1000 ohm - 1uF with 15 averages -
https://dl.dropbox.com/u/3090951/7samples2.png
[20:24:12] <PCW> You will always need a analog input filter with arcs at they will have significant high frequency noise (which will show up as baseline wander/aliased waveforms to a sampled A-D)
[20:24:20] <jdh> It's lovely, I'm sure.
[20:24:38] <tjb1> Its a colorful picture!
[20:24:45] <tjb1> PCW im moving up to 2k 1uf now
[20:26:20] <jdh> PCW: how long have you been doing stuff like this?
[20:26:44] <PCW> too long
[20:26:56] <jdh> heh... you rock!
[20:29:48] <jdh> you probably have to select 'delimited' and comma for the text-to-column
[20:30:05] <PCW> bye all
[20:39:33] <tjb1> yeah that delay killed it
[20:39:38] <tjb1> looks like a sine wave now
[20:42:57] <jdh> you need to sample at least 2x the freq of whatever it is you want to see.
[20:45:33] <tjb1> Im just testing to get the best for the output I need
[20:47:08] <tjb1> I think my consumables are getting shot
[20:47:49] <tjb1> The same filter, same cutting parameters only difference is averging 8 samples -
https://dl.dropbox.com/u/3090951/1660k.png
[20:52:20] <jdh> this will be the feedback for the controller?
[20:52:26] <tjb1> grr im stuck in russian linuxcnc forum
[20:52:45] <tjb1> arduino is going to eventually take the vfd output and use it to find out what voltage I want the arc to be
[21:05:07] <tjb1> im going to attempt part of that now
[21:05:12] <tjb1> is || or in arduino?
[21:06:00] <jdh> yes
[21:07:47] <tjb1> so this would work "if(realVoltage =< 118.2 && realVoltage >= 117.8)"
[21:07:52] <tjb1> erm...
[21:08:06] <tjb1> = <
[21:08:42] <tjb1> I guess its <= and >=
[21:08:51] <jdh> yep
[21:09:36] <t12> have you scoped pre/post filter while running it
[21:09:45] <jdh> I'd swap the order, and put parens in :)
[21:09:50] <t12> to see if the filter is really doing what you want
[21:10:00] <tjb1> Yes
[21:10:16] <tjb1> https://dl.dropbox.com/u/3090951/filtervsnot.jpg
[21:10:29] <tjb1> Green = filtered, red = not filtered
[21:10:32] <ReadError> what are you using to graph?
[21:10:40] <tjb1> Thats excel ReadError
[21:10:43] <ReadError> oh
[21:10:47] <tjb1> :)
[21:10:52] <tjb1> Nothin fancy
[21:11:03] <ReadError> what you tryin to measure
[21:11:06] <ReadError> voltage i know
[21:11:08] <ReadError> but what for
[21:11:36] <ReadError> oh nm
[21:11:37] <ReadError> tjb1> arduino is going to eventually take the vfd output and use it to find out what voltage I want the arc to be
[21:12:25] <tjb1> Im going to use "M3 S117" to tell the arduino I want an arc voltage of 117 using the G540 VFD output and then the arduino will sample the arc voltage, compare it to the VFD setting and tell Linuxcnc to move torch up or down
[21:14:01] <t12> can you servo directly on the voltage?
[21:14:37] <jdh> seems like that woudl be better.
[21:14:39] <t12> like cant you just manipulate the voltage output directly
[21:15:20] <tjb1> I have to read what the torch is at
[21:15:27] <t12> of course but
[21:15:39] <t12> rather than manipulating arc voltage with torch height
[21:15:43] <t12> why not just change the voltage
[21:15:57] <tom3p> the voltage is teh result of the position, not an input
[21:15:57] <tjb1> Not something you do with plasma
[21:16:08] <jdh> just use the arc voltage as feedback, let linuxcnc control the height
[21:16:12] <tjb1> Only control you have is the amperage and the height
[21:16:14] <tom3p> proximity/position
[21:16:20] <t12> ah ok
[21:16:21] <tjb1> jdh, I dont have a fancy mesa card :)
[21:16:36] <jdh> you can't just use PID?
[21:16:56] <tjb1> how would I get the voltage into the computer?
[21:17:04] <jdh> serial
[21:18:11] <jdh> http://axis.unpy.net/01198594294
[21:18:29] <ink-> hm
[21:18:40] <ink-> do any of you have any good resources for PID loop tuning?
[21:18:57] <jdh> there are lots of books on the subject.
[21:19:11] <ink-> any recommendations?
[21:19:20] <ReadError> wiki has a good article
[21:19:25] <jdh> heh, it is a voluminous subject.
[21:19:31] <ink-> right :)
[21:19:38] <tjb1> Oh you want me to use the arduino to send voltage over serial?
[21:19:40] <jdh> jt & pcw always seem to have the answers.
[21:19:41] <ink-> hm maybe I'll stop by the engineering library and poke around
[21:19:44] <jdh> jtb1: right.
[21:19:56] <tjb1> That requires interfacing arduino with computer
[21:20:04] <tjb1> I wont need it connected to computer once im done
[21:20:12] <jdh> tjb1: then let linuxcnc close the loops... since it is driving the motor anyway
[21:20:22] <tjb1> What loops?
[21:20:42] <jdh> the closed loop torch height controller you would be implementing.
[21:21:40] <tjb1> What should I do differently?
[21:22:00] <jdh> what moves your torch up/down now?
[21:23:02] <tjb1> linuxcnc
[21:24:41] <jdh> how does it know where to move it?
[21:24:43] <tjb1> http://pastebin.com/Qe55XSAR
[21:24:52] <tjb1> Those leds will soon be inputs to linuxcnc
[21:25:08] <tjb1> if input torch up is high…it will move torch up until it goes low
[21:25:16] <jdh> heh
[21:25:24] <jdh> interesting baud rate
[21:26:37] <tjb1> do you understand it now?
[21:26:44] <jdh> <tjb1> I wont need it connected to computer once im done
[21:27:03] <tjb1> Its connected to the BOB
[21:27:13] <tjb1> not usb and require the arduino run on the computer
[21:27:53] <ReadError> what kind of modulation do you have on the plasma cutters power unit?
[21:27:56] <ReadError> is it a smooth dial
[21:28:08] <tjb1> you select amperage
[21:28:08] <jdh> it doesn't use the arduino dev stuff. just a comp
[21:28:09] <ReadError> or predetermined clicks ?
[21:28:42] <ReadError> im too accident prone for anything like a plasma cutter ;/
[21:28:51] <ReadError> ive sustained 2 hand injuries in the past 2 weeks
[21:28:59] <ReadError> i put my finger into a belt sander
[21:29:02] <ReadError> and drilled into my hand
[21:31:06] <tjb1> Its a pot ReadError
[21:32:31] <jdh> your led stuff looks backwards to me.
[21:32:56] <tjb1> It is :/
[21:33:01] <tjb1> Thats why it wouldnt work haha
[21:33:40] <tjb1> All I have to do is get those 2 outputs into my c10 bob and into linuxcnc
[21:34:08] <ReadError> tjb1, cant you use a servo to modulate the pot ?
[21:34:16] <jdh> now change the order of the last case to satisfy my coding OCD.
[21:34:40] <tjb1> ReadError: changing the amperage isnt used to change the voltage on plasma
[21:35:06] <tjb1> jdh: How can I get my outputs into this board?
[21:35:23] <jdh> I'd put optos between them.
[21:35:26] <tjb1> Set the inputs as pulled up and pull up the outputs on the arduino and set them as low?
[21:35:39] <tjb1> Set them as low to activate them
[21:35:46] <jdh> either way works fine.
[21:35:53] <tjb1> then share the ground between arduino and the bob?
[21:36:02] <jdh> I'd use optos :)
[21:36:06] <tjb1> why
[21:36:12] <jdh> to isolate them
[21:36:39] <tjb1> that BOB isolated them from computer
[21:36:56] <jdh> how much is the BOB?
[21:37:23] <tjb1> $28
[21:37:34] <jdh> guess it doesn't matter then:)
[21:37:58] <ReadError> which BOB ?
[21:38:02] <jdh> pull-ups on the arduino, drive them low-active
[21:38:18] <tjb1> then pullups on the BOB inputs?
[21:38:25] <jdh> yeah
[21:38:42] <jdh> broken wire will make them both on.
[21:39:40] <tjb1> I can pull the inputs low
[21:39:45] <tjb1> and drive them high ;)
[21:40:08] <jdh> seems likely to be 'safer' that way
[21:40:18] <tjb1> The shared ground would allow that to work right
[21:40:32] <jdh> you have to share it either way
[21:41:40] <jdh> seem like the arduino hal component would make this a simple servo axis
[21:44:36] <jdh> I assume the hardcoded numbers will be replaced with values read from the vfd output?
[21:46:12] <tjb1> yes
[21:46:20] <tjb1> no resistors needed on arduino side if I pull inputs low
[21:46:26] <tjb1> or do I need to pull the outputs low also?
[21:47:26] <jdh> pull the bob low
[21:47:50] <tjb1> Thats it
[21:48:40] <jdh> 5v arduino?
[21:48:45] <tjb1> yes
[21:52:46] <tjb1> jdh: Any comments on this?
http://pastebin.com/fp8F6jdA
[21:54:20] <jdh> heh, nope.
[21:56:42] <tjb1> if Im pulling high
[21:57:02] <tjb1> The pin would not need -not at the end correct
[21:57:05] <tjb1> in hal
[21:57:11] <jdh> yep.
[21:57:21] <jdh> he has a mesa reference in there also
[21:57:37] <tjb1> where?
[21:57:54] <jdh> net thc-pos-cmd thcud.z-pos-out => hm2_[HOSTMOT2](BOARD).0.stepgen.02.position-cmd
[21:58:21] <tjb1> erm...
[21:58:38] <tjb1> what do I need to replace that with?
[21:58:52] <jdh> whatever your stepgen.02 is using
[21:59:51] <tjb1> my hal file says "net zpos-cmd axis.2.motor-pos-cmd => stepgen.2.position-cmd
[22:03:15] <jdh> I would guess the right side of yours if adequate
[22:04:20] <tjb1> So I need "net thc-pos-cmd thcud.z-pos-out => stepgen.2.position-cmd
[22:04:57] <jdh> that woudl be my guess (for that line)
[22:05:21] <tjb1> What is emcmot?
[22:05:22] <tjb1> net emcmot.02.pos-cmd thc.z-pos-in <= axis.2.motor-pos-cmd
[22:06:01] <jdh> motion controller
[22:06:35] <tjb1> So that can stay the same then
[22:06:43] <jdh> yeah
[22:08:36] <tjb1> anything else you see needs changed?
[22:11:05] <jdh> nope, but I am not that knowledgable.
[22:13:22] <tjb1> Do I need to add line 24 and 25 to my hal file?
[22:13:52] <jdh> those are your inputs from teh arduino
[22:14:06] <tjb1> I know, do those need to be in the HAL file?
[22:14:16] <jdh> no, that is 30,31
[22:14:40] <tjb1> So 30 and 31 dont need to be in the HAL file either?
[22:15:10] <jdh> 30,31 do
[22:15:49] <tjb1> can those stay in the comp file?
[22:15:52] <jdh> 24,25 are just text descriptions
[22:16:06] <jdh> they can stay there also.
[22:16:20] <jdh> everything in side the triple quotes is just text.
[22:17:21] <jdh> the HAL lines in 30-37 have to go in your HAL
[22:18:56] <tjb1> the two files I need at the end…what extension?
[22:19:42] <jdh> none
[22:19:54] <tjb1> just save as M101 and M102
[22:19:55] <jdh> might need to chmod +x them though?
[22:20:15] <tjb1> I dont even know what that means
[22:20:26] <jdh> tells the OS they are executable files
[22:20:50] <jdh> dunno how linuxcnc calls them though. Might not be needed (but probably is)
[22:21:43] <tjb1> I remember you have to open properties and make them executable
[22:22:01] <jdh> or just type "chmod +x M101"
[22:24:22] <tjb1> I need to restart my laptop…brb
[22:24:48] <jdh> I need to go read an old-fashioned paper book. Good luck.
[22:25:08] <tjb1> 1 sec!
[22:26:49] <tjb1> computer about died...
[22:29:53] <tjb1> Im trying to paste the new ones but I have to make an account...
[22:32:40] <tjb1> jdh:
http://pastebin.com/b5AvNqNX
[22:32:56] <tjb1> http://pastebin.com/zbDWks6K
[22:32:59] <tjb1> do those look ok? :D
[22:51:05] <tjb1> Im trying to install a comp file and its telling me "sudo: comp: command not found" after running "sudo comp --install thcud.comp"
[22:52:59] <ReadError> i dont have comp installed on my linuxcnc box
[22:53:53] <tjb1> so I need to install that?
[22:56:44] <tjb1> found it
[23:27:49] <tjb1> r00t4rd3d: Help me :(
[23:34:08] <r00t4rd3d> with what
[23:37:01] <tjb1> Well to install a comp file I had to install emc2-dev
[23:37:09] <tjb1> which I guess erased linuxcnc and put emc2 on
[23:39:20] <r00t4rd3d> build emc2-dev from source but dont make install
[23:40:43] <tjb1> how do I install 2.5 back again?
[23:41:03] <tjb1> Is it safe to just install 2.5 again and build emc2-dev?
[23:41:04] <r00t4rd3d> update it
[23:41:45] <tjb1> link to do so?
[23:41:51] <r00t4rd3d> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?UpdatingTo2.5
[23:42:51] <tjb1> Just follow 10.4 then build emc2-dev
[23:45:46] <tjb1> I guess I can just install linuxcnc-dev at the same time
[23:46:19] <r00t4rd3d> probably
[23:46:36] <r00t4rd3d> Ive never installed the dev stuff
[23:46:57] <tjb1> I did get the linuxcnc files working on the emc2 lol
[23:47:07] <tjb1> the comp file didnt work though but the table worked again...
[23:49:34] <tjb1> thanks a bunch r00t4rd3d
[23:49:37] <tjb1> Back on 2.5 :)
[23:50:12] <r00t4rd3d> now update it to 2.5.1
[23:50:17] <tjb1> it did
[23:50:21] <tjb1> im having another problem
[23:51:09] <tjb1> want to try again? :D
[23:51:28] <tjb1> value 'parport.1.pin-12-in' invalid for bit
[23:51:31] <tjb1> setp failed
[23:52:12] <r00t4rd3d> that is beyond me
[23:52:51] <tjb1> I think it was because it was missing "net" before it
[23:52:57] <r00t4rd3d> hal and mesa and all that junk
[23:53:22] <tjb1> Signal name must not be the same as a pin. did you omit the signal name?
[23:54:22] <r00t4rd3d> i just built my first Android kernel :D
[23:54:52] <tjb1> nice
[23:56:02] <tjb1> oh my…i figured it out
[23:56:08] <tjb1> Been one hell of a day haha
[23:57:20] <r00t4rd3d> what are you trying to do now?
[23:57:40] <tjb1> if i add these mcodes
[23:57:44] <tjb1> the THC should work now :D
[23:58:15] <r00t4rd3d> try better pot
[23:59:05] <tjb1> nothin :(
